Researchers from the Max Planck Institute have developed a mass spec-based workflow for quantifying the secreted proteome, or secretome, of activated immune cells.
The workflow, detailed in a study published this week in Science, offers a method for secretome analysis that is both antibody-free and streamlined enough to be used for applied biological and clinical research, said Felix Meissner, first author on the paper and researcher in the group of Max Planck professor Matthias Mann, whose lab led the effort.
Secreted proteins are key components of cellular signaling and a potentially rich source of disease biomarkers, but their low abundance makes them difficult targets for proteomic analysis, especially given the complexity of the serum-containing media in which the cells of interest are usually grown.
As Meissner told ProteoMonitor, "The issue [with secretome] analysis is that when you start going into clinical samples or serum samples, you have the high abundance proteins [in the serum-containing media] that [expand] your dynamic range." This, he noted, prevents the mass spectrometer from seeing the low-abundance secreted proteins of interest.
This being the case, researchers have typically used antibody-based methods for studying these proteins, but because of constraints on antibody availability and multiplexing, such approaches can detect only a limited portion of the secretome.
To get around these problems, the Max Planck researchers devised a mass spec workflow wherein they used reduced concentrations of serum in their media, thereby reducing the complexity of the sample. Even with this approach, they still ran into interference from high-abundance serum proteins, Meissner noted, but, he said, they were also able to quantify a significant portion of the lower-abundance secreted proteins present.
In total, the team quantified 775 secreted proteins, including 52 known cytokines, achieving picogram levels of sensitivity. In addition to the use of reduced-serum media, advances in mass spec technology played a significant role in the study's success, Meissner said.
The researchers used a Thermo Fisher Scientific Q Exactive machine, which, he said, offered the sequencing speed necessary to reach into the low-abundance portions of the secretome.
In preliminary experiments, Meissner used a Thermo Scientific Orbitrap Velos, with which he was able to quantify roughly half as many proteins as he and his colleagues ultimately did with the newer instrument.
"Moving to the Q Exactive was a really big step," he said, adding that the machine's capabilities made the group's mass spec approach competitive with the immunoassays typically used for detection of low-abundance secreted proteins like cytokines.
More important than the workflow's ability to quantify a broad complement of secreted proteins, however, was its speed, Meissner said. He noted that in their study, the researchers performed their analyses using two-hour LC gradients and single mass spec runs, collecting data on more than 120 samples in roughly two weeks.
"If you spend, [for example], five times as much [mass spec] measuring time or fractionate the samples, then in principle the study would also be possible on the Velos," he said. "But then it's not that interesting."
With the timescale demonstrated in the Science study, the workflow "is something that you can start to really think about [using for] biological experiments," Meissner said. "We're now going in a direction... [where] we have the ability to start analyzing big bunches of samples."
Jeroen Krijgsveld, a researcher at the European Molecular Biology Laboratory, who with colleagues recently authored a separate study presenting a click chemistry-based approach to secretome analysis (PM 10/5/2012), suggested to ProteoMonitor, however, that the study's use of reduced-serum media could prove problematic.
Such an approach "is a bit tricky, because these macrophages need serum to be happy," he said, noting that reducing the amount of serum in the growth media could change the composition of the secretome.
The Max Planck group "spent a lot of time finding the optimal conditions where they could minimize the serum without damaging the cells," said Krijgsveld, who was not a part of the study. However, he said, there is still a possibility the change in serum levels could alter the proteins secreted.
"It's a question mark," he said.
Krijgsveld's team approached their secretome analysis via an enrichment strategy that labeled proteins with azidohomoalanine, or AHA, an azide-containing analog of the amino acid methionine. Once secreted, the azide-labeled proteins could then be captured by alkyne-activated beads via click chemistry, allowing for pull-down of secreted proteins prior to mass spec analysis.
In their study, published in September in Nature Biotechnology, the EMBL researchers analyzed PC3 cells – prostatic adenocarcinoma cells initiated from a bone metastasis — and WPMY-1 cells – myofibroblast stromal cells from healthy prostate – identifying a total of 1,136 secreted proteins and reliably quantifying 684.
Krijgsveld said that he and his colleagues are now using the method to examine cell to cell interactions, looking in particular at the secretions of cancer and immune cells when exposed to each other.
Meissner acknowledged the concerns Krijgsveld cited regarding the reduced-serum media, noting that it "is something that has to be taken seriously" and that for each biological system studied researchers would have to first test to determine the technique's suitability.
In the Science study, the researchers were looking at non-dividing primary cells with low metabolisms, he said, noting that this made them amenable to the method. "We tested how our cells would behave with different concentrations of serum. ... We titered out [the serum] very carefully."
Meissner also suggested that presenting the data as fold-change comparisons between sets of cells served as another check on the workflow.
"One of the important aspects is that we always present our data in a ratiometric way, so we always compare two conditions, treated and untreated," he said. "So even in the worst case ... starvation induced autophagy [of the cells under study], then that effect would be read out in both conditions."
Meissner said that one advantage of his team's technique compared to enrichment-based methods is that it can work with relatively small amounts of cells.
"Usually when you are using enrichment techniques you will need several million cells" due to sample losses during the enrichment procedure, he said, adding that he and his colleagues were able to use as few as 150,000 cells with their approach.
The researchers are now using the method to further investigate some of their initial findings, Meissner said, including a number of secreted proteins they identified that were not expected to have an extracellular function.
"There is a lot of biological follow-up" to be done, he said.